It seems the code and comment have inconsistent encoding currently as of 4.11.0.3. Newly added low-end electromagnetic processes (G4ANSTOecpssrKxsModel.hh etc) on comment header line have special latin character – in windows-1252 code page (central europe and us?):
// - S. Bakr et al. (2021) NIM B, 507:11–19.
The sequence of bytes is not valid in utf-8.
If /source-charset:windows-1252 is specified in cmake cxx flags on windows, the lines with °°° will evaluate to °°°.
Suggested fix would be chage the – to a normal ascii -, conssistent with the next line:
Same thing in G4INCLXXInterface.cc string literal description:
The Liège Intranuclear Cascade (INCL++)
In utf-8, this got evaluated to
The Li�ge Intranuclear Cascade (INCL++)
Complete code:
void G4INCLXXInterface::ModelDescription(std::ostream& outFile) const {
outFile
<< "The Liège Intranuclear Cascade (INCL++) is a model for reactions induced\n"
<< "by nucleons, pions and light ion on any nucleus. The reaction is\n"
<< "described as an avalanche of binary nucleon-nucleon collisions, which can\n"
<< "lead to the emission of energetic particles and to the formation of an\n"
<< "excited thermalised nucleus (remnant). The de-excitation of the remnant is\n"
<< "outside the scope of INCL++ and is typically described by another model.\n\n"
<< "INCL++ has been reasonably well tested for nucleon (~50 MeV to ~15 GeV),\n"
<< "pion (idem) and light-ion projectiles (up to A=18, ~10A MeV to 1A GeV).\n"
<< "Most tests involved target nuclei close to the stability valley, with\n"
<< "numbers between 4 and 250.\n\n"
<< "Reference: D. Mancusi et al., Phys. Rev. C90 (2014) 054602\n\n";
}